Machine Learning Regression 4

Question 1

In simple linear regression, what does the slope coefficient represent?

Accepted Answer

The change in Y for a one-unit increase in X

Answer

The slope coefficient β₁ represents the expected change in the dependent variable Y for each one-unit increase in the predictor X.

Question 2

Which loss function is more robust to outliers in regression?

Accepted Answer

Mean Absolute Error (MAE)

Answer

MAE is more robust to outliers because it uses absolute differences, whereas MSE squares errors and thus amplifies large residuals.

Question 3

What is 'regularization' primarily designed to prevent in regression models?

Accepted Answer

Overfitting

Answer

Regularization penalizes model complexity to prevent overfitting, encouraging simpler models that generalize better.

Question 4

In multiple linear regression with p predictors, what is the adjusted R² designed to correct for?

Accepted Answer

The increase in R² from adding irrelevant predictors

Answer

Adjusted R² penalizes for the number of predictors, preventing the illusion that adding variables always improves the model.

Question 5

What is the key advantage of quantile regression over ordinary least squares?

Accepted Answer

It models specific quantiles of the conditional distribution, not just the mean

Answer

Quantile regression estimates any conditional quantile (e.g., median, 90th percentile), providing a richer picture of the response distribution.

Machine Learning Practice Test